There has been considerable research in the use of near-infrared optical spectroscopy (NIRS) as a means for real-time in-vivo measurements of tissue optical properties which contain information on tissue structure and function. In the 600-1000 nm spectral region in particular, tissue is scattering dominated and the strongest molecular absorbers in tissue are oxygenated and deoxygenated hemoglobin, water, and lipids. The highly diffusive photons probe a large sample volume, providing macroscopically averaged absorption and scattering properties at depths up to a few centimeters.
The amount of light reflected or transmitted from tissues is due to a complex combination of absorption, scattering and (typically very weak) fluorescence. In order to measure any of these optical properties of a given sample, one has to first separate/isolate the absorption effects from scattering effects. Possession of this capability is enabling for a wide range of medical (diagnostic, therapeutic monitoring, cosmetic) and non-medical applications (material inspection, visualization, photo-realistic rendering, agricultural inspection, chemical slurry and powder analysis).
NIR techniques (though actually not limited to the NIR spectral range) combine experimental measurements and model-based data analysis to quantitatively measure the bulk absorption (μa) and scattering (μs′) properties of the tissue. Once μa and μs′ are known at a variety of wavelengths, the concentration of the various molecular absorbers can be determined.
Several techniques have been developed over the last decade to measure tissue properties in vivo, and they can be broadly grouped into two categories: (1) photon migration techniques and (2) optical biopsy techniques. Most instruments of these types rely on a fiber optic contact probe measurement so that the source-detection geometry is well defined. The geometry allows for the quantitative measurement of absorption and scattering properties of the tissue, but it is limited to a single, small area. Photon migration instruments usually use source-detector separations of a few centimeters, resulting in spatial resolutions on the order of one centimeter, such that μa and μs′ can be determined for thick tissue, such as breast, brain and muscle. Optical biopsy techniques usually use source-detector distances on the order of 100's of microns, thus they interrogate a smaller spatial scale that is typically on the order of one millimeter.
For many medical diagnostic applications, there is need for techniques that combine some of the physiological information that photon migration and optical biopsy provide, but have a wide field, non-contact imaging capability. Multispectral imaging systems that use a camera with a tunable spectral light source (or spectral detection filters) have been used in this capacity. There is a fundamental issue, however, on the inability of camera systems to distinguish between light that is absorbed by the tissue and light that is scattered. Imaging systems that use full-field illumination (i.e. flash photography) cannot differentiate between the two effects and assumptions are made in order to provide “quantitative” biochemical analysis. In practice, this deficiency results in qualitative analysis that depicts relative concentration changes within an image.
A more detailed discussion of these techniques is provided in Cuccia. Modulated Imaging: A Spatial Frequency Domain Imaging Method for Wide-field Spectroscopy and Tomography of Turbid Media, Ph.D. Dissertation, University of California, Irvine, Dept. of Biomedical Engineering (“Cuccia, Modulated Imaging”); and Cuccia, et al., Quantitation and mapping of tissue optical properties using modulated imaging, J Biomed Opt 14 (2), 024012 (2009) (“Cuccia, Quantitation and mapping”).
Due to the deficiencies of prior techniques, a technique and technology platform, referred to as “Modulated Imaging” (MI), was developed. The key aspect of this type of imaging is that the absorption and scattering components are separated and used to evaluate tissue structure and calculate quantitative biochemical maps. The MI method uses structured light projection and camera-based detection in order to obtain quantitative measurements of:
1. sub-surface tissue optical properties, including:                a. tissue absorption due to:                    i. endogenous chromophores such as oxy- and deoxy-hemoglobin, water, lipids, melanin, bilirubin, porphyrins, etc. and            ii. exogenous dyes such as indocyanine green, methylene blue, synthetic agents, etc.                        b. tissue fluorescence/phosphorescence due to subsequent remission of light after absorption from a molecule above        c. tissue scattering, (microscopic refraction) including both scattering magnitude and direction, due to:                    i. cellular structures such as nuclei, mitochondria, cell membranes,            ii. extracellular structures, such as collagen            iii. exogenous agents                        
2. surface profile information (profilometry)
A detailed description of the MI method including spatial frequency domain imaging (SFDI) measurement, calibration, and analysis has been previously reported in Cuccia, Quantitation and mapping, and U.S. Pat. No. 6,958,815, which are incorporated herein by reference.
From an apparatus perspective, an innovative aspect of MI/SFDI is its combination of a camera and a structured light projection system that allows one to reconstruct quantitative maps in 2D or 3D of tissue optical properties. Structured light illumination, also commonly referred to as spatially structured illumination, includes, among other, such illumination as sinusoidal illumination and periodic illumination. Generally, the structured illumination patterns give multiple “views” into the tissue and reveal the contrast between various structures and optical properties that would otherwise be obscured or mixed together. The system is typically non-contact, allowing for easy use in applications including surgical guidance where the tissue of interest can be interrogated without contamination.
From a method perspective, the camera images can be analyzed in a variety of ways to extract this quantitative information. The most common embodiment is “spatial frequency domain” analysis, involving either 1) processing a single Fourier-transform of the images, or 2) by directly manipulating a series of images under multiple structured illumination conditions—typically a spatial sine wave at various spatial phases. A strong benefit to methods that use approach 2) above is that they lend themselves more readily toward recovering high-resolution maps (in 2D or 3D) of the recovered properties, thus allowing for spatially-resolving structures and/or determining the depth of various structures/layers/etc.
Another innovation aspect of MI is the combination of simultaneous profile measurements along with tissue optical property determination. Profilometry is commonplace in areas such as machine vision and cosmetic dermatology.
As stated above, MI has the unique capability of spatially resolving optical absorption and scattering parameters, allowing wide-field quantitative mapping of tissue optical properties with the use of spatially-modulated illumination. FIG. 1 shows the configuration of a laboratory-grade system 10. Light from a halogen lamp 11 is expanded by a condenser 12 onto a spatial light modulator (SLM) 15. The current system uses a Digital Micromirror Device (DMD) from Texas Instruments which is a 1024×768 mirror array that can generate and project arbitrary grayscale patterns. Such patterns are directed through a projector lens 16 and reflected off a mirror 17 to the surface of the tissue T and the diffusely reflected light is then recorded by a digital CCD camera 19. In the laboratory instrument, a filter wheel 13 was used to interrogate a discrete number of wavelengths. Instead of a filter wheel, a tunable filter or tunable spectral source can be used to interrogate a discrete number of wavelengths. Crossed linear polarizers 14 and 18 can be introduced into the source and detection light paths to remove specular reflectance. The SLM 15, camera 19 and spectral device are synchronized with a computer and/or trigger board, enabling fast acquisition of a series of patterns at various spatial frequencies. A turbid reflectance standard (such as a TiO2-based silicone phantom) can be used to calibrate the source intensity and to correct for spatial non-uniformities in both the illumination and imaging systems.
Periodic illumination patterns of various spatial frequencies are projected over a large (many cm2) area of a sample. Typically, sine-wave illumination patterns are used. The reflected image captured by the camera differs from the illumination patterns due to the optical property characteristics of the sample. The demodulation of these spatially-modulated waves characterizes the sample modulation transfer function (MTF), which embodies the optical property information of the tissue.
For example, the tissue can be illuminated with a spatial pattern of the form:
                    S        =                                            S              0                        2                    ⁡                      [                          1              +                                                M                  o                                ⁢                                  cos                  ⁡                                      (                                                                  2                        ⁢                                                  πf                          x                                                                    +                      α                                        )                                                                        ]                                              (        1        )            where S0, M0, fx and α are the illumination source intensity, modulation depth, spatial frequency, and spatial phase, respectively. The diffusely reflected intensity, I, is a sum of the spatially-varying (AC) and spatially-constant (DC) components of the illumination signal. These AC and DC spatial components do not relate to other uses of the terms AC and DC, such as the AC and DC components of electrical signals, or the AC and DC temporal components, for example those delineated in Sevick-Muraca U.S. Pat. No. 5,865,754. The underlying physics, detection schemes, analysis methods and mathematical models aimed at characterizing these AC and DC spatial components are all distinct from other uses of these terms.
The top row of images, shown in FIG. 2, show the images obtained for illumination patterns at four spatial frequencies (with only 1 phase of each frequency shown). The AC component of the reflected intensity, IAC, can be modeled as:IAC=MAC(x,fx)·cos(2πfx+α)
Here, MAC(x,fx) represents the amplitude of the reflected photon density “standing wave” at frequency fx. Note that MAC can be a function of position, x. To obtain MAC(x,fx), a simple time domain amplitude demodulation method is employed, illuminating a sinusoid pattern three times at the same spatial frequency, with phase offsets α=0, 2/3π and 4/3π radians. MAC(x,fx) can then be calculated algebraically at each spatial location, xi, by:MAC(x,fx)=[(I1−I2)2+(I2−I3)+(I3−I1)]1/2  (3)The spatially-varying DC amplitude, MDC(x), can be calculated using:MDC(x,fx)=[I1+I2+I3]/3  (4)where I1, I2, and I3 represent the IAC image values at each location with shifted spatial phases.
Finally, measurement of a reference turbid phantom of known optical properties allows model-based calibration for the source intensity, S0, and therefore conversion of MAC and MDC to calibrated diffuse reflectance, RAC and RDC, respectively. Once the AC and DC components of the reflectivity are determined, a “White Monte Carlo” (WMC) method is used to provide accurate and rapid models of predicting light transport over a wide range of reflectivities. At each wavelength, the spatial-frequency-dependent diffuse reflectance is fitted to WMC forward predictions for every pixel in the image and obtain the μa and μs′ optical properties, as shown at the bottom of FIG. 2. This can be performed with a rapid two-frequency lookup table using a minimal 3-phase, single frequency image set (by demodulating and averaging the images to obtain AC and DC amplitude maps, respectively). This simple algorithm can easily be implemented for real-time processing and/or implementation on camera FPGA hardware. Alternatively, this analysis could be performed via other predictive, statistical, or heuristic models.
By mapping the absorption coefficient at multiple wavelengths, quantitative spectroscopy of tissue can optionally be performed. The result is a 3D data cube with an absorption spectrum at each spatial location. Knowledge of the extinction coefficients of the tissue chromophores (e.g. oxy- and deoxy-hemoglobin, lipids, water, etc) allows these spectra to be fitted to a linear Beer-Lambert absorption model and determine the quantitative concentrations of each chromophore.
Any of the aforementioned point detection systems, measurements, and analyses could be further spatially multiplexed to yield 1D, 2D, or 3D spatial representations of the turbid medium optical properties and/or structures. From a hardware standpoint, this would include multiple copies of a previously described detector setup, or an optical relay or scanning system to relay detector information from various locations on the sample.
As described above, the generally regarded innovative aspect of MI is the combination of a camera (2D light sensor) and a structured illumination system (2D projector) to enable the measurement and 2D/3D mapping of optical properties and tissue structures. Although this system can be constructed with consumer-grade electronics, it nevertheless requires a certain level of cost and complexity due to the presence of a 2D sensor. For example, when the method is extended for spectroscopy (measurement of multiple wavelengths), it adds significant system complexity and/or measurement time constraints, requiring either serial single-wavelength measurements or bulky and expensive multi-spectral imaging systems. In addition, although combination/integration with measurement methods that use time-modulation of light is also theoretically possible (in addition to spatial structuring or spatial modulation of light), this has never been feasible or desirable as it requires expensive, bulky, and low-fidelity time gating systems for cameras.
Thus, it is desirable to provide a less costly and complex system to analyze the optical properties and structures of turbid media.
It should be noted that elements of similar structures or functions are generally represented by like reference numerals for illustrative purpose throughout the figures. It should also be noted that the figures are only intended to facilitate the description of the preferred embodiments.